init

app.py

@@ -55,7 +55,7 @@ hubert = torch.hub.load("bshall/hubert:main", "hubert_soft")
 
 _ = utils.load_checkpoint("nyarumodel.pth", net_g_ms, None)
 
-def vc_fn(input_audio,vc_transform):
+def vc_fn(sid, input_audio,vc_transform):
     if input_audio is None:
         return "You need to upload an audio", None
     sampling_rate, audio = input_audio
@@ -78,7 +78,7 @@ def vc_fn(input_audio,vc_transform):
     sf.write("temp.wav", audio22050, 22050)
     pitch = transcribe("temp.wav", soft.shape[0], vc_transform)
     pitch = torch.LongTensor(pitch).unsqueeze(0)
-    sid = torch.LongTensor([0])
+    sid = torch.LongTensor([0]) if sid == "猫雷" else torch.LongTensor([1])
     stn_tst = torch.FloatTensor(soft)
     with torch.no_grad():
         x_tst = stn_tst.unsqueeze(0)
@@ -89,16 +89,20 @@ def vc_fn(input_audio,vc_transform):
     return "Success", (hps_ms.data.sampling_rate, audio)
 
 
-
 app = gr.Blocks()
 with app:
     with gr.Tabs():
         with gr.TabItem("Basic"):
-            vc_input3 = gr.Audio(label="Input Audio (30s limitation)")
-            vc_transform = gr.Number(label="变调数量",value=1)
-            vc_submit = gr.Button("Convert", variant="primary")
+            gr.Markdown(value="""本模型相比与前一个模型，音质和音准方面有一定的提升，但是低音音域目前存在较大问题。
+            目前能够唱的最低音为#G3(207hz) 低于该音会当场爆炸（之前的模型只是会跑调），
+            因此请不要让这个模型唱男声的音高，请使用变调功能将音域移动至207hz以上。
+            """)
+            sid = gr.Dropdown(label="音色",choices=['猫雷',"？？？"], value="猫雷")
+            vc_input3 = gr.Audio(label="上传音频（长度小于45秒）")
+            vc_transform = gr.Number(label="变调（整数，可以正负，半音数量，升高八度就是12）",value=0)
+            vc_submit = gr.Button("转换", variant="primary")
             vc_output1 = gr.Textbox(label="Output Message")
             vc_output2 = gr.Audio(label="Output Audio")
-        vc_submit.click(vc_fn, [ vc_input3,vc_transform], [vc_output1, vc_output2])
+        vc_submit.click(vc_fn, [sid,  vc_input3, vc_transform], [vc_output1, vc_output2])
 
     app.launch()